K ' K K - ACS Publications

peated with 30 ml of 50% aqueous ethanol. The combined aqueous ethanol solutions were cooled to give 2.3 g of 2. When the oily portion insoluble in aq...
2 downloads 0 Views 673KB Size
MARCH1967

ALKYLATION OF ACTIVEHYDROGEN COMPOUNDS

peated with 30 ml of 50% aqueous ethanol. The combined aqueous ethanol solutions were cooled to give 2.3 g of 2. When the oily portion insoluble in aqueous ethanol was allowed to stand at room temperature, 2.9 g of phenyl disulfide, mp 60°, separated. The yield was 90%. Reaction of p-Nitrobenzenethiol with 1 .-Five grams (0.029 mole) of 1 was added to a solution of 9 g (0.058 mole) of p-nitrobenzenethiol in 100 m! of absolute ethanol. After refluxing for 8 hr, the reaction mixture was allowed to stand at room temperature, and the crystals were separated by filtration and recrystallized from ethanol to give 8 g (90%) of p-nitrophenyl disulfide, mp 18G182.5'. Reaction of o-Aminobenzenethiol with 1 .-A solution of 2.5 g (0.02 mole) of o-aminobenzenethiol and 3.5 g (0.02 mole) of 1 in 50 ml of anhydrous benzene was refluxed for 4 hr. After cooling, the crystals which deposited were separated by filtration and recrystallized from methanol to give 2.3 g (677,) of oaminophenyl disulfide, mp 93'. Reaction of 2-Naphthalenethiol with 1 .-To a solution of 2.3 g (0.014 mole) of 2-naphthalenethiol in 50 ml of anhydrous chloroform was added 2.5 g (0.014 mole) of 1. After refluxing for 5 hr, the reaction solution was evaporated under reduced pressure to remove the chloroform. The residue was recrystallized from ethanol to give 2 g of 2-naphthyl disulfide, mp 132'. The yield was 88%). Reaction of 2-Mercaptobenzothiazole with 1 .-Five grams (0.029 mole) of 1 was added to a solution of 9.7 g (0.058 mole) of 2-mercaptobenzothiazole in 100 ml of anhydrous benzene (or

729

ethanol), and the mixture was refluxed for 30 min. After cooling, the crystals which separated were filtered and recrystallized from benzene or ethanol give 2,2'-dithiobisbenzothiazole, mp 181°, in quantitative yield.

Registry No.-1, 1972-28-7 ; ethanol, 64-17-5; isopropyl alcohol, 67-63-0 ; cyclohexanol, 108-93-0; benzyl alcohol, 100-51-6; p-nitrobenzyl alcohol, 619-73-8; a-methylbenzyl alcohol, 98-85-1 ; benzhydrol, 91-01-0; dodecanol, 112-53-8; 1,2-propanediol, 57-55-6 ; semicarbazone derivative of 1,7429-48-3 ; 2,4-dinitrophenylhydrazone derivative of 1, 7429-49-4 ; hydrazobenzene, 122-66-7; p-anisidine, 104-94-9; ethanethiol, 75-08-1 ; 2-propanethiol, 75-33-2 ; 2-propene-1-thiol, 870-23-5 ; 1-dodecanethiol, 112-55-0; benzenethiol, 108-98-5; p-nitrobenzenethiol, 1849-36-1; o-aminobenzenethiol, 13707-5 ; 2-naphthalenethiol, 91-60-1 ; 2-mercaptobenzothiazole, 149-30-4. Acknowledgment.-We are grateful to Dr. Chikayoshi Nagata for his suggestions, assistance, and guidance during the course of this investigation. We express also our appreciation to Dr. Tomoichiro Akiba, Director of this laboratory, for his encouragement.

Alkylation of Active Hydrogen Compounds by N-Vinylamides R. A. HICKNER,C. I. JUDD,' AND W. W. BAKKE Chemicals Department Research Laboratory, The Dow Chemical Company, Midland, Michigan .@740 Received October 1.6,1966 Active hydrogen compounds such as amides, urethans, sulfonamides, thiols, and alcohols are alkylated by N-vinylamides, urethans, or sulfonamides under acidic conditions in high yields.

The acid-catalyzed alkylation of 2-pyrrolidinone (1) by N-vinyl-2-pyrrolidinone (2) has been reported by Breitenbach.2vs More recently a similar reaction has

7"

+

0

~ C H = C H ,

--t

1

~

-

0

0

~ CH3

H

-

0

2

N

~ CH3

-n 0 NCH=CHz K 0

+

0 3 1 been reported between 2-oxazolidinone and N-vinyl2-0xazolidinone.~ As part of a developmental program on derivatives of K-vinyl-5-methyl-2-oxa~olidinone5~6product from 2 and 4 had mp 91-95', mmp 92-96");? however, the identity of the two products was supported by elemental analysis, infrared, and nmr. n r 1 n n O NH + 0 NCH=CHz 0 N-CH-N 0 Since cyclic amides and carbamates reacted, it K K ' K seemed reasonable to assume that open-chain amides or 0 0 O CHs 0 urethans would react similarly. Indeed, acetamide, benzamide, or acrylamide were alkylated in excellent (3), we have investigated acid-catalyzed reactions of yields by either N-vinyl-5-methyl-2-oxazolidinone (3) 3 and related derivatives. or N-vinyl-2-pyrrolidinone (2). The double bond If the original mechanism suggested by Breitenbach is correct, one should obtain the same product from ( 3 3 3 7 - 7 1 and 3 or from 2 and 5-methyl-2-oxazolidinone (4). 3 RCONHa + 0 N-CH-NHCOR We found that this is, indeed, the case. Mixture meltK I ing points were not completely definitive (the melting 0 CH3 point of the product from 1 and 3 was 94-98'; the R = Me, phenyl, vinyl

K

-

+

(1) Lakeside Laboratories, Milwaukee Wis. (2) J. W. Breitenbach, F. Galinovsky, H. Nesvabda, and E. Wolf, Monatsh., 81, 580 (1956). (3) J. W.Hreitenbach, J . Polymer Scz., 18, 949 (1957). (4) A. Kutner, J . Ore. Chem., 96, 3495 (1961). (5) W.E.Walles. W. F. Tousignant, and T. Houtman, Jr., U. S. Patent 2,891,058(1959). (6) W.W.Bakke, 1J. S. Patent 2,905,690(1959).

of acrylamide is retained producing an N-substituted acrylamide. The new monomers polymerized readily with typical radical catalysts, but all efforts produced (7) Since the product contains two asymmetric centers, it will consist of a pair of diastereoisomers. Selective loss of small amounts of one isomer could

readily account for the differences.

HICKNER, JUDD, AND BAKKE

730

only insoluble gels. The gellation of the polymers is likely due to trace amounts of N,N'-ethylidinediacrylamide. The formation of this could be due to traces of acetaldehyde present in the N-vinyl compound, since it was found that all samples of 3 contained acetaldehyde a t levels up to several hundred parts per million. Alternately this could be explained m the basis of a "transalkylation" reaction. Studies presented later 3+RCONHz

"

+

C

CH3 5 t RCONHz

& n O,,,N-YH-NHCOR

" --+

+H

Z

5

RCONHYHNHCOR CH3

in this work indicate that transalkylation is apparently not a serious side reaction. However, we can not discount the possible formation of trace amounts of the ethylidine compound by this route. Frequently less than 1-2% of' cross-linking agent is sufficient to cause gellation in polymers. The behavior of sulfonamides in this reaction was investigated next because of their formal resemblance to amides. All sulfonamides investigated gave the expected products in yields over 90%. The high yields with sulfonamides was probably due to a combination of high reactivity and low solubility of the products in the reaction medium. The high reactivity of the sulfonamides was also demonstrated by the ready reaction of benzenesulfonamide with the relatively hindered N-vinyl-N-methylbenzenesulfonamide.

I n view of the reactivity of amides and sulfonamides it was anticipated that imides would behave similarly. However, we were unsuccessful in affecting reaction between succinimide or phthalimide and 3. Mercaptans are another class of compounds possessing an electron pair that should undergo alkylation. We found that mercaptans are probably the most reactive class of compounds studied. Mercaptans may also undergo spontaneous free-radical addition to olefins ; however, free-radical addition would occur a t the terminal carbon atom.81~Attachment of the alkylthio group to the same carbon as the nitrogen was proved by hydrolysis of the adduct with aqueous acid in the presence of 2,4-dinitrophenylhydrazineto give acetaldehyde (isolated as the 2,4-dinitrophenylhydrazone), thus indicating the ionic nature of the reaction. Infrared analysis also indicated that under acid conditions attachment of the alkylthio group always occurred a t the nonterminal carbon. The acidcatalyzed products from N-vinyl-2-oxazolidinones all ( 8 ) F. W. Staoey and J. F. Harris, Jr., Org. Reactions, 19, 164 (1963).

(9) M.F. Shostakovskii, F. P. Sidel'kovskoya, and F. 9. Kolodkin, Bull. Acad. Science USSR, 155 (1962).

VOL. 32

showed a strong, sharp bond at 8.05 p while 3-(2-alkylthioethyl)-2-oxazolidinones prepared from the 3-(2chloroethyl)-2-oxazolidinones and sodium mercapt i d e ~ all ~ ~had ' ~ a strong sharp band a t 7.95 p which was absent in the acid-catalyzed products. Comparison of the free-radical adduct from 2 and 1-octanethiol CH37-7 0 NCHzCHzCl

K 0

+ RSNa

C H s n 0 N-CHZCHZSR

-+

K 0

by glpc also indicated the absence of any significant amounts of free-radical product in the acid-catalyzed reaction." In order to test our previous hypothesis regarding the transalkylation reaction, N-vinyl-2-pyrrolidinone was treated with a molar excess of 1-octane- or 1pentanethiol. No 2-pyrrolidinone or dithio acetal was detected by glpc. The resistance of dithio acetals to V-Z:;SR

+

RSH

---t

s

NH

0

+

RS-CH-SR I CH3

6

hydrolysis12in contrast to the relative ease of hydrolysis of the 1-[ (l-alkylthio)ethyl]-2-pyrrolidinonessuggests that if transalkylation is a significant factor in these reactions, it should be particularly favorable in the thiol case. However, when 6 was treated with concentrated hydrochloric acid, the dithio acetal was isolated in 43%. The presence of 2-pyrrolidinone was shown by dPC* N-Vinylsuccinimide was found to be unreactive toward thiols, in accord with our observations of the nonreactivity of imides in this system (see above). N-Vinyl-N-methylbenzenesulfonamidereacted readily with the thiols as would be predicted by the reactivity of sulfonamides with the N-vinylamides studied. It occurred to us that vinyl ethers might alkylate amides in the same manner as the vinylamides; however, the only product isolated was the corresponding ethylidine bis compound. l 3 Apparently 7 undergoes

protonation as it is formed to produce the same intermediate which is obtained by protonating the N-vinyl derivative. This intermediate may then react further (10) R. A. Hickner, U. S. Patent 3,188,317(1965). (11) The apparently purposeless switch from the oxaaolidinone to pyrrolidinone series in several places in this paper was dictated by physical properties of the products. The former series was used where lower solubility and higher melting points was desired while the latter series was studied when higher volatility was desired. (12) (a) A. Schoenberg and W. Asker, J . Chem. Soc., 604 (1946);(b) E. Campaigne, "Organic Sulfur Compounds," Val. 1, N. Kharasch, Ed., Pergamon Press Inc., New York, N. Y., 1961,pp 134-145. (13) R.A. Hiokner, W.W. Bakke, and C. I . Judd, U. 5. Patent 3,072,652 (1963).

MARCH1967

ALKYLATION OF ACTIVEHYDROGEN CONPOUNDS

731

were dissolved or suspended in benzene. Anhydrous hydrogen chloride was bubbled in until the temperature began to rise. External cooling was applied if necessary to prevent the temperature from rising above 50". Most products precipitated readily from benzene with the exception of the thiol adducts. The solid products were collected by vacuum filtration and in CH3-i some cases recrystallized from ethanol. The thiol adducts 0 N-CHCH, K were distilled through a flash still. The products prepared, 0 physical properties, and analyses are listed in Tables I and 11. All products, except the thiol adducts, developed the odor of with 5-methyl-2-oxazolidinone. To test this hypothesis, acetaldehyde when left exposed in air. N-vinyl-2-pyrrolidinone was treated with 1-butanol in Reaction of N-Vinyl-2-pyrrolidinone (2) with 5-Methyl-2oxazolidinone (4).-A solution of 20.2 g (0.2 mole) of 4 in 50 ml the presence of anhydrous hydrogen chloride. That of benzene was acidified to pH 1 with anhydrous HCl and 22.2 g (0.2 mole) of 2 was added dropwise while maintaining the temperature below 40. Stirring was continued an additional 0.5 hr. The precipitate was dissolved in methylene chloride and the solution was neutralized with anhydrous ammonia, and filtered to remove the salt. The solvent was removed at reduced pressure. 8 The solid collected by vacuum filtration and washed with etherbenzene ( 5 : l ) . The solid weighed 35.5 g (84.17,). After a the expected reaction had occurred was indicated by the second washing as described above the product melted at 87.591". disappearance of the infrared absorption a t 2.9 and Anal. Calcd for CloHlsNzOa:C, 56.35; H, 8.00; N, 13.1. 6.15 p for the hydroxyl and double bond, respectively, Found: C, 56.60; H, 7.68; N, 13.07. and appearance of a strong ether band at 9.1 p . HowA small sample of the above product recrystallized from 2B ever, attempts to distil 8 resulted in disproportionation absolute ethanol melted at 91-95'. to a complex mixture of butanol, dibutyl acetal, and Reaction of N-Vmyl-5-methyl-2-oxazolidinone(3) with 2Pyrrolidinone (l).-A solution of 16.8 g (0.2 mole) of 1 in 50 other unidentified products.14 When crude 8 was ml of benzene was treated wit,h 25.4 g (0.2 mole) of 3 as in the treated with 1-octanethiol, I-[ (octylthio)ethyl]-2-pyrabove example. The crude product was dissolved in methylene rolidinone was produced as evidenced by glpc comparichloride, neutralized with anhydrous ammonia, and filtered. son with the product from X-vinyl-2-pyrrolidinone and The crude product was triturated with ether-benzene ( 5 :1) and dried to give 31 g (74%) of product melting at 93-96'. 1-octanethiol. Anal. Calcd for CloH16r\;203:C, 56.35; H, 8.00; N, 13.1. Found: C, 56.65; H, 7.72; N, 13.35. The above crude product after retrituration with etherbenzene melted at 94-98". A mixture melting point of this product with that from 2 and 4 was 91.5-98'. The infrared spectra of the two products were essentially identical in solution I n summary, we can conclude that (1) under acidic in CCld or CSZor as mulls. The nmr spectra showed no sensible differences. conditions the reactions of N-vinylamides, N-vinylPolymerization of Products from Acrylamide and N-Vinyl-2sulfonamides, vinyl ethers, and N-( 1-alkoxyalky1)A.-A solupyrrolidinone or N-Vinyl-5-methyl-2-oxazolidinone. amides differ in degree rather than kind, and (2) most tion of 15 g of freshly prepared product from N-vinyl-5-methyl-2compounds containing an NH adjacent to a single oxazolidinone, 15 g of 3, and 0.35 g of a,a'-azobisisobutyronitrile was heated at reflux for 2.5 hr. The reaction mixture a t this carbonyl or sulfonyl group or containing an SH should point was a gel which could not be stirred. undergo acid-catalyzed alkylation by N-vinyl comB .-A solution of 30 g of product from N-vinyl-2-pyrrolidinone pounds in which the nitrogen is adjacent to a single and 300 mg of azobisisobutyronitrile in 70 ml of 507, aqueous carbonyl or sulfonyl group. methanol was heated at reflux for 10 min a t which time an insoluble gel had formed. The product was also insoluble in water, 95% ethanol, acetone, or chlorobenzene. Experimental Section C.-A solution of 20 g of product from X-vinyl-5-methyl-2oxazolidinone, 1 g of concentrated ammonium hydroxide, and Equipment.-All melting points were determined on a Fisher0.1 g of 307, aqueous hydrogen peroxide was heated to 70" Johns block. An F & M Model 500 gas-liquid partition chrounder a nitrogen atmosphere. After a few minutes an insoluble matograph was used with a 2 ft X 0.25 in. SE 30 on Chromosorb gel had formed. W column, and a helium flow rate of 80 cc/min. The gas-liquid Attempted Reaction of N-Vinyl-5-methyl-2-oxazolidinone (3) partition chromatograph was programmed at 15'/min from an and Phtha1imide.-A suspension of 73.6 g (0.5 mole) of phthalinitial temperature of 100". imide in 63.5 g (0.5 mole) of 3 and 200 ml of benzene was treated Materials.-The thiols, amides, imides, and sulfonamides were with anhydrous hydrogen chloride. No temperature rise ocpurchased from Eastman. 2-Pyrrolidinone and N-vinylS-pyrcurred. After stirring for 1 hr, the mixture was filtered to give rolidinone from General Aniline and Film Co., were redistilled 71 g (96.5%) of material which did not depress the melting point through an 8-in. Vigreux column. N-Vinylsuccinimide was of phthalimide. purchased from Monomer-Polymer Laboratories and n-butyl Attempted Reaction of N-Vinyl-5-methyl-2-oxazolidinone and ether was obtained from Union Carbide Corp. 5-Methyl-2Succinimide.-Succinimide (49.5 g, 0.5 mole) suspended in a oxazolidinone was prepared by the method of Close,15 bp 121solution of 63.5 g (0.5 mole) of 3 and 200 ml of benzene was acidi124' (2 mm), n% 1.4595. The N-vinyl-2-oxazolidinones were prepared by direct vinylation:b N-vinyl-5-methyl-2-oxazolidinone fied with anhydrous hydrogen chloride. A small temperature rise occwred, apparently owing to heat of solution. After stirring bp 88-89' (1.5 mmj, T Z ~ D1.4752; N-vinyl-5-ethyl-2-oxazolifor 1 hr the solid (4,5 g) was collected by filtration. A mixture dinone, bp 84-86' (0.8 mm), 1 2 % ~1.4762. N-Vinyl-N-methylmelting point with succinimide was undepressed. benzenesulfonamide was prepared by the method of Cairns and Ultraviolet-Catalyzed Addition of 1-Octanethiol to N-Vinyl-2Sauer,16bp 139-140" ( 4 mm), n 2 61.5464 ~ (lit.lBn% 1.5468). pyrrolidinone (2) .--il 600-ml beaker was placed in a water bath General Procedure for Addition Reactions.-One mole of the through which cold water was circulated. The bath was placed N-vinyl compound and 1 mole of the active hydrogen compound on a magnetic stirrer and a stirring bar was used to stir the contents of the beaker. The beaker was charged with 43.8 g (0.3 (14) M.F. Shostakovskii, F. P. Sidel'kovskaya, and M. G . Zelenskays, mole) of 1-octanethiol and 33.3 g (0.3 mole) of 2 was added dropBull. Acad. Sci. USSR,488 (1959). wise while irradiating with a G.E. UA-2 lamp. After irradiating (15) W.J . Close, J. A m . Chem. Soc., 75, 95 (1951). for 6 hr the crude product was charged to a flash still. After (16) T.L. Cairns and J. C. Sauer, J. Ore. Chem., 20, 627 (1955).

+

HICKNER, JUDD,AND BAHHE

732

VOL.

32

TABLE I ALKYLATION OF AMIDE-TYPE COMPOUNDS Yield, N-Vinyl compd

MP, 'C4

Active H oompd

%

-Calcd, Formula

C

%-

-Found, C

H

%H

N-Vinyl-2-pyrrolidinone Acr ylamide b 90 CeHiiNzOz 59.38 7.70 59.37 7.58 N-Vinyl-2-pyrrolidinone Benzenesulfonamide 143-145 95 CizHisNzOsS 53.72 5.97 54.16 6.34 5-Methyl-2-oxazolidinone 91-95c N-Vinyl-2-pyrrolidinone 84 C10&6N203 56.35 8.00 56.60 7.68 Acetamide N-Vinyl-5-methyl-2-oxazolidinone 145-148 50 C8Hi4NzOa 51.48 8.28 51.57 7.58 N-Vinyl-5-methyl-2-oxazolidinone Acrylamide b 53 CeHiiNzOs 54.55 7.06 55.23 7.32 p-Toluenesulfonamide N-Vinyl-5-methyl-2-oxazolidinone 91 C l & ~ N ~ O ~ S 52.38 6.04 143-146 52.76 5.98 Benzamide N-Vinyl-5-methyl-2-oxazolidinone 93 CiaHisNzOa 62.90 6.45 156-159 62.84 6.43 N-Vinyl-5-methyl-2-oxazolidinone 2-Pyrrolidinone 74 CioHieNzOs 56.35 8.00 94-98 56.63 7.72 N-Vinyl-5-methyl-2-oxazolidinone Benzenesulfonamide 145-148 93 CizHisNzOS 50.70 5.63 50.38 5.55 N-Vinyl-N-methylbenzenesulfonamide Benzenesulfonamide 58 CisH18N20~2 50.83 5.08 133-136 50.92 5.01 Many of these compounds are mixtures of two diasteriomers so that recrystallization can concentrate isomers and result in different melting points. * Polymerizes. See discussion regarding these products. 0

TABLEI1 ALKYLATION OF THIOLS -Found,

H

C

N-Vinyl-2-pyrrolidinone N-Vinyl-2-pyrrolidinone N-Vinyl-5-me thyl-2-oxazolidinone N-Vinyl-5-methyl-2-oxazolidinone N-Vinyl-5-methyl-2-oxazolidinone N-Vinyl-N-methylbenzenesulf onamide N-Vinyl-N-me thylbenzenesulfonamide N-Vinyl-5-e thyl-2-oxazolidinone a Not distilled.

H

11.17 10.48 10.70 6.36 10.57

69.10 65.02 65.93 60.46 64.90

11.23 10.59 10.97 6.43 10.15

63.42

9.38 63.73

9.33

CisHisNOzSzCl 54.02 CizHz3NOB 58.73

5.06 54.16 9.45 58.84

4.90 9.20

-Cslcd, C

Bp, OC (mm)

nzsD

%

n-Dodecyl n-Octyln-DodecylBenzenen-Octyl-

155-165 (0.2) 144 (0.5) >160 (0.5) a a

1.4844 1.4896 1.4752 1.5570 1.4778

88 87 87 75 87

CisHasNOS CirHzTNOS CiaHssNOB CizHiaNOiS Ci4Hz~N02S

68.95 65.35 65.70 60.72 65.30

n-Dodecyl-

a

1.5090

84

CvHsrNOzSz

a 1.5870 13Ck140 (0.15) 1.4804

98 30

Thiol

p-Chlorobenz yl3-Methylbutyl-

removing 3.6 g of lights boiling up to 148' (0.1 mm), two product cuts were collected: cut A, bp 147-150' (0.3-0.5 mm), n% 1.4958; cut B, bp 147" (0.3 mm), 12% 1.4963. Anal. Calcd for C14H2,NOS (cut B): C, 65.30; H, 10.57. Found: C, 64.90; H, 10.30. The nmr of EL 20% CDCla solution supported the structure. The retention time on glpc (Ti = loo", 15"/min) was 12.9 min while the acid-catalyzed adduct (see below) had a retention time of 12.6 min. A glpc of the mixed products resolved the two components readily. Structure Proof of Adduct from N-Vinyl-2-pyrrolidinone and 1-Dodecanethio1.-A solution of 3.13 g (0.01 mole) of the title product, 1.98 g (0.01 mole) of 2,4-dinitrophenylhydrazine,15 ml of water, 50 ml of 2B absolute ethanol, and 5 ml of sulfuric acid was heated at reflux for 30 min. On cooling 1.30 g (65%) of the product crystallized out, which melted a t 146-147' after trituration with ethanol. A mixture melting point with acetaldehyde2,4-dinitrophe1iylhydrazone was undepressed. Structure Proof of Adduct from N-Vinyl-5-methyl-2-oxazolidinone and 1-Dodecanethio1.-A solution of 3.27 g (0.01 mole) of adduct was reacted as above to give 1.30 g (65%) of 2,4-dinitrophenylhydrazone which did not depress the melting point of an authentic sample. Preparation of 1,l-Bis(octy1thio)ethane (Dioctyl Thial of Acetaldehyde) .-1-Octanethiol (29.2 g, 0.2 mole) was acidified with anhydrous HC1 and 10.0 g (0.1 mole) of butyl vinyl ether was added dropwise. Occasional cooling was applied to keep the temperature below 50". Stirring was continued for an additional 0.5 hr and the crude product was transferred to a flash still. A forecut (7.0 g) was taken up to a head temperature of 135" (0.21 mm) followed by 28.5 g (89.5%) of product, bp ~ 135-142' (0.2-0.4 mm), n s 1.4814. Anal. Calcd for ClsH&32: C, 67.90; H, 12.03; S, 20.13. Found: C, 67.92; H, 11.69; S, 20.00. Acid-Catalyzed Addition of 1-Octanethiol to 2.-A solution of 58.4 g (0.4 mole) of 1-octanethiol in 25 ml of methylene chloride was acidified to pH 1with anhydrous hydrogen chloride and 44.4 g (0.4 mole) of 2 added dropwise with cooling to keep the temperature below 48". A 20-ml sample was neutralized with anhydrous ammonia and filtered, and the solvent was removed. This product eluted faster on the glpc than the radical product (see above).

%-

%-

Yield, N-Vinyl compd

Formula

Another 20-ml sample was pipetted out (pH still 1) and 5 ml of 1-octanethiol was added. Samples were taken a t intervals over a several-hour period and submitted to glpc analysis. No dithio acetal was detected. [The above experiment was repeated using 1-pentanethiol. The single product was eluted in 8.0 min (Ti = loo', 15'/min)]. A 13-g sample of the adduct was treated slowly with 25 ml of concentrated HC1. Initially the reaction was clear, but became cloudy within a few minutes. After standing for 10 min, the two-phase system was extracted with chloroform and the solution was dried with drierite, and evaporated. The crude product (3.6 g, 43%) was shown by glpc to be identical with the dithio acetal, 12% 1.4808. Anal. Calcd for C18Hs~Sz: C, 67.90; H, 12.03; S, 20.13. Found: C, 68.40; H, 11.55; S, 19.65. Glpc analysis of the aqueous phase showed the presence of 2-pyrrolidinone. Reaction of 1-Butanol with N-Vinyl-2-pyrrolidinone.-A solution was prepared of 111 g (1.0 mole) of N-vinyl-2-pyrrolidinone and 74 g (1.0 mole) of 1-butanol. The infrared spectrum had a strong band a t 6.15 p for the double bond and a strong hydroxyl band. The flask was cooled to 12" with an ice bath and anhydrous HCl was sparged in gently until the temperature began to rise. The temperature rose rapidly to 45'. The cooling bath was removed and stirring was continued for 1 additional hr. The mixture was then stirred for 0.5 hr with anhydrous sodium carbonate. The infrared spectrum indicated nearly complete absence of double bond and hydroxyl. Distillation through an 8-in. Vigreux column gave a liquid fraction of 110 ml which boiled at 63-98' (3.0 mm) (most boiling at 9698"). Infrared showed a moderate hydroxyl band at 2.9, a moderate C=C band at 6.18, and a broad ether band a t 9.1 p . The presence of dibutyl acetal was supported by comparison with an authentic sample. A solid residue weighing 43 g remained in the pot. I n Situ Reaction of I-( l-Butoxyethyl)-2-pyrrolidinone with I-Octanethio1.-1-Butanol (29.6 g, 0.4 mole) was acidified with anhydrous hydrogen chloride and 44.4 g (0.4 mole) of N-vinyl-2pyrrolidinone wtts added dropwise over a 10-min period. The temperature rose to 45". Upon cooling to room temperature, the infrared spectrum showed no double-bond absorption at 6.1 fi and only a slight hydroxyl absorption.

ACRIDIZINIUM IONCHEMISTRY. VI

MARCH1967

1-Octanethiol (58.4 g, 0.4 mole) was added dropwise during 6 min to the above product which was still acid whereupon the temperature rose to 43'. After stirring for 1 hr, anhydrous sodium carbonate and a dropperful of concentrated ",OH were added. The vapor phase chromatograph indicated the product to be a mixture of I-butanol and I-[(1-octy1thio)ethylJ2-pyrrolidinone. Distillation of a 65-g sample through a flash still gave a forecut of butanol and 46.7 g (9ly0) of the octylthio compound, bp 144-146' (0.5 mm), n% 1.4892. Anal. Calcd for C14H2,NOS: C, 65.30; H, 10.48. Found: C, 64.95; H, 10.15. The retention time on the glpc was identical with that of the acid-catalyzed, product of 1-octanethiol to 2. Attempted Reaction of N-Vinyl uccinimide with I-Dodecanethiol.-A solution of 10.9 g (0.1 mole) N-vinylsuccinimide and 20.2 g (0.1 mole) of I-dodecanethiol in 50 ml of methylene chloride a t 21' was treated with 5 drops of concentrated sulfuric acid. After several minutes when no apparent reaction had occurred, and additional 5 drops of sulfuric acid was added. After standing for an additional 45 min the infrared spectrum of the solution was identical with the starting solution.

733

Registry No.-C9H14N202, 7594-58-3 ; C1zHl&zO3S, 7594-59-4 ; CsH14Nz03, 7594-61-8; CgH14N203, 759462-9; CHHISNZO~S, 7594-63-0; C I ~ H I ~ N ~ 7594-64-1 O~, ; CIOHIBNZO~, 7594-60-7 ; CIZHI~J'JZO~S 7594-66-3 ; C16H18Nz04SzJ 7594-67-4 ; C I ~ H ~ ~ N O 5681-84-5; S, C14Hz7NOS, 7594-69-6 ; CISH~SNOZS, 5681-87-8 ; CizHuNOzS, 7594-72-1 S, ; Cz1H,,N02S, 75945681-86-7 ; C U H ~ ~ N O Z 73-2 ; C I ~ H D N ~ Z S Z 7594-74-3 C~, ; CizH&OzS, 568185-6; C1sH&z,7594-76-5. Acknowledgment.-The authors are indebted to L. B. Westover and Dr. J. C. Gavan for infrared spectra and to Dr. J. P. Heeschen for nmr analyses. The elemental analyses were performed by DOW'SSpecial Services Laboratory. Helpful discussions with Professors hlartin Stiles and Jack Crump are acknowledged.

Acridizinium Ion Chemistry. VI.'

Reaction with Bases. 11'

C. K. BRADSHER AND J. P. SHERER Department of Chemistry, Duke University, Durham, North Carolina Received September 6, 1966 I n basic solution the acridishiurn ion under goes ring opening and can give base-catalyzed condensation of the type to be expected if %(>formylbenzyl)pyridine were present. The easily prepared oxime of 2-(2-fomylbenzy1)pyridine has been converted to 2-( 2-cyanobenzyl)pyridine which proved to be a convenient starting material for the preparation of acridizinium and benzo [b]quinolizin-6-one derivatives.

I n what was apparently the first paper to describe a quinolizinium derivative, Schneider and Schroeter3 showed that coralyn (8-methyl-2,3,10,11-tetramethoxybenz[a]acridieinium, I) cation, in the presence of a base, underwent ring opening to. afford a methyl ketone (11). By analogy it would be expected that addiOCH3

2

9

1

11

.

10

\

\

IV

111

0

V

I

CH3

IL

tion of base to the acridieinium ion (111) might yield 2- (2-formylbenzy1)pyridine (V), probably via the pseudo-base (IV). Frost and Saylor4 have invoked this equilibrium to explain the change in mode of polarographic reduction and the nature of the ultraviolet absorption spectrum in this system with increasing pH. It was reported earlier6ssthat addition of base to an (1) For the preceding oommunication of this series, Bee C. K. Bradsher and J. D. Turner, J . 0 7 ~ .Chem., 81, 565 (1900). (2) This investigation waa supported by Public Health Service Research Grant No. CA-05509 of the National Cancer Institute. (3) W. Schneider and K. Schroeter, Ber., 68, 1459 (1920). (4) J. G. Frost and J. H. Saylor, Rec. Trau. C h h . , 89, 828 (1963). (5) A. Richards and T. 8. Stevens, J . Chem. Soc., 3067 (1958).

aqueous solution of acridizinium ion (111) leads to the precipitation of a red or brown powder. Although this powder has not yet been purified, evidence has been adduced (Experimental Section) which suggests that the material is a mixture of the aldehyde (V) and the pseudo-base (IV). Evidence for the existence of the aldehyde (V) in basic alcohol solution was afforded by sodium borohydride reduction to afford the corresponding benzyl alcohol, 2-(2-hydroxymethylbenzyl)pyridine,which was isolated as the methoperchlorate. Proton magnetic resonance measurements support the assigned structure. The most convincing chemical evidence for the presence of the formylbenzylpyridine (V) in an acridizinium solution that has been made basic was afforded by the formation of an oxime (VIIIa, 87%) or a semicarbazone (VIIId, 76%). The proton magnetic resonance spectra of the oximes (VIa-c) made it quite clear that the ring had opened in the indicated manner. On (6) C. K. Bradsher and J. H. Jones, J . Am. Chem. Soc., 81, 1938 (19592